opennn::Correlation Struct Reference

Result of a correlation analysis: model parameters, fit quality, and the method/form used. More...

#include <correlations.h>

Public Types
enum class	Method { Pearson , Spearman }
	Underlying coefficient family used to compute r. More...
enum class	Form {   Identity , Sigmoid , Logarithmic , Exponential ,   Power }
	Functional form fitted to the data before computing the coefficient. More...

Public Member Functions
	Correlation ()
void	set_perfect ()
	Sets the correlation to a perfect linear relationship (a=0, b=1, r=1).
void	print () const
	Prints the correlation parameters and method to stdout.

Public Attributes
float	a = NAN
float	b = NAN
float	r = NAN
float	lower_confidence = NAN
float	upper_confidence = NAN
Method	method = Method::Pearson
Form	form = Form::Identity

Detailed Description

Result of a correlation analysis: model parameters, fit quality, and the method/form used.

Member Enumeration Documentation

Form

enum class opennn::Correlation::Form

strong

Functional form fitted to the data before computing the coefficient.

Enumerator
Identity
Sigmoid
Logarithmic
Exponential
Power

Method

enum class opennn::Correlation::Method

strong

Underlying coefficient family used to compute r.

Enumerator
Pearson
Spearman

Constructor & Destructor Documentation

Correlation()

opennn::Correlation::Correlation ( )

inline

Member Function Documentation

print()

void opennn::Correlation::print

(

)

const

Prints the correlation parameters and method to stdout.

set_perfect()

void opennn::Correlation::set_perfect

(

)

Sets the correlation to a perfect linear relationship (a=0, b=1, r=1).

Member Data Documentation

a

float opennn::Correlation::a = NAN

b

float opennn::Correlation::b = NAN

form

Form opennn::Correlation::form = Form::Identity

lower_confidence

float opennn::Correlation::lower_confidence = NAN

method

Method opennn::Correlation::method = Method::Pearson

r

float opennn::Correlation::r = NAN

upper_confidence

float opennn::Correlation::upper_confidence = NAN